Analysis of Different Image Processing Techniques for Classification and Detection of Cancer Cells

Analysis of Different Image Processing Techniques for Classification and Detection of Cancer Cells

Bukhtawar Elahi (National University of Sciences and Technology, Pakistan), Maria Kanwal (National University of Sciences and Technology, Pakistan) and Sana Elahi (Dr. A. Q. Khan Institute of Computer Sciences and Information Technology, Pakistan)
Copyright: © 2020 |Pages: 25
DOI: 10.4018/978-1-7998-2521-0.ch008


This chapter gives an analysis of various methodologies for detecting cancer cells through image processing techniques. The challenges during such detections are over-segmentation and computational complexities. Therefore, the algorithms dealing with such problems are analyzed in this chapter. In these algorithms, a watershed and setting up threshold are helpful to overcome segmentation issues. A support vector machine is discussed to detect subtypes of pneumoconiosis for disjointing segments of lungs. For finding lung cancer cells, a segmentation weighted fuzzy probabilistic-based clustering has been used. Multiple variants of thresholding along with classifiers are proposed to detect lungs and liver cancer. Other than that, noise-removal, feature extraction and watershed are used to detect breast cancer. For leukemia, a bimodal thresholding over enhanced images of cytoplasm and nuclei regions has been discussed. kNN classifier, k-mean clustering, and feed-forward neural networks have also been discussed. Results from these techniques vary from 60%-100% depending on the proposed methodology.
Chapter Preview

1. Introduction

Cancer cells have the power to affect the other healthy cells and the blood vessels surrounding these cancerous cells. This area forms an environment known as microenvironment. The cancerous cells may invade the other cells and spread to influence the other faraway body parts. These cancerous cells form a mass that increases in size to form tumors, lumps or masses. The detection of masses of cancer cells can be detected by using different techniques. Depending on the nature of their invasion, cancer is categorized as malignant or benign as shown in Figure 1.

Figure 1.

Benign VS malignant tumor


Benign tumor does not invade or spread to other body parts though its size can be increased. After the removal of this type of tumor it does not affect again but malignant tumor can come back to affect the other body parts. Benign tumor is not life threatening, however, malignant tumor is fatal if it reaches to brain. Malignant tumor invades other surrounding parts but small part of the originally located tumor dislodge and enters into the circulatory system or lymphatic system where it goes to the distant body parts to make them cancerous. Cancer cells use the immune system for their growth and can mask themselves from immune cell and hence immune system becomes unable to remove cancer cells from the body.

There are different types of cancer and their categorization depends on the organ or tissue where it is present and types of cells. Leukemia is the cancer of blood or blood forming cells. Lymphoma is the cancer of lymphatic system. Carcinoma is the cancer of lining that surrounds the skin, digestive tract and lungs. Sarcoma is the cancer of mesodermal cells such as bone, vessels and blood. Melanoma is the cancer of melanocytes containing melanin pigment. Some other cancers include breast cancer, brain and spinal cord cancer etc (Geetha & Selvi, 2015). Lung cancer involves the cancer of trachea, bronchi and lung tissue involving alveoli. Smoking, exposure to radon gas, silica and asbestos, air pollution, lung diseases such as tuberculosis and family history are the main causes of the lung cancer. Lung cancer stages are different depending upon the cancer size and its severity. The stage of cancer can be diagnosed with highly precise screening tests.

Skin cancer is the mostly widespread cancer. It starts from the cells that line the superficial layer of the skin. It is localized type of cancer but most destructive in terms of damaging the skin tissues. Basically, there are various kinds of skin cancer that includes squamous cell carcinoma (first kind of skin cancer), basal cell carcinoma (second kind of skin cancer) and melanoma (the third type). Exposure to ultraviolet rays, immunologically suppressed body, family history and exposure to ionizing rays are the main causes of skin cancer. Blood cancer affecting the blood cells number and their function, bone marrow and lymphatic system. Mainly it affects the bone marrow (tissues) where the blood cells are developed. Leukemia, lymphoma and melanoma are the main types of blood cancer. Chemotherapy and radiotherapy could be used as the treatment of his cancer.

Brain tumor originates in the brain and it is a localized tumor known as benign tumor. If not diagnosed early, it can be life threatening. There are some common types of brain tumor that includes gliomas and astrocytic tumors and meningeal tumor. Seizures, vision problem, memory loss problem, speaking, hearing, balancing difficulties are some symptoms of this cancer. Chemotherapy and surgically removal of this tumor are treatments for brain tumors.


2. Background

Cancerous cells exhibit a large variety of attributes due to which radiologists have been misinterpreted the results obtained through screening programs. Computer aided classification and detection algorithms are being developed for better interpretation of the results. Reduction of false results, effective and authentic results could now obtain by using detection system (Geetha al., 2015).

Key Terms in this Chapter

Image Processing: To extract useful information from images different algorithms and actions are performed on an image.

Leukemia: Cancer that effect blood cells.

KNN: K nearest neighbors is a classification algorithm that classifies the object based on k nearest neighbors.

Tumor: A tumor is a mass of tissue that is framed by an aggregation of anomalous cells.

CT Scan: Computed tomography (CT) scan uses data from several X-rays and combine them to get the image of particular part of that object.

Segmentation: In image segmentation, the users select only that area of image that is of their interest. In this process the image is divided into multiple sections or structures of the interest, so that each section has characteristic similarity.

Melanoma: Type of skin cancer.

MRI: Magnetic resonance imaging (MRI) uses radio waves and magnetic fields to get images of the tissues and organs of the body.

Cancer Identification: Different approaches are used for diagnosis and identification of cancer. Different test and procedures are used to diagnose different cancers.

Fuzzy Control System: Fuzzy logic-based control system that takes on continuous values (between 0 to 1) as input.

Lymphoma: Cancer that starts in the lymphocytes.

Complete Chapter List

Search this Book: